We have been studying crystallins, the abundant water-soluble proteins responsible for the optical properties of the eye lens, as a model of tissue-specific gene expression. In FY2003 we have continued to examine the mouse small heat shock protein (shsp)/ alphaB-crystallin gene locus. A shsp/alphaB-crystallin-related gene (MKBP/HSPB2), which was derived by duplication of the shsp/alphaB-crystallin gene during evolution, is present 1 kb upstream of the mouse shsp/alphaB-crystallin gene and is transcribed in the opposite direction. The MKBP/HSPB2 gene is expressed in muscle and heart, but not in lens. Last fiscal year we reported that the intergenic alphaB-crystallin enhancer affects the shsp/alphaB-crystallin promoter in an orientation-specific manner, and has surprisingly little influence on the MKBP/HSPB2 promoter. This year we established that a GC-rich sequence between the MKBP/HSPB2 gene and the enhancer can act as a silencer and serves to insulate the MKBP/HSPB2 promoter from the enhancer. We have continued our collaborative project with Dr. Eviatar Nevo (University of Haifa, Israel) on the blind mole rat. This rodent develops an eye during embryogenesis that regresses and has a degenerate lens fragment at best in the adult. We showed earlier that the alphaB-crystallin promoter/enhancer of the mouse and blind mole rat are similar but not identical. Luciferase reporter transgenes driven by either a mole rat or a mouse alphaB-crystallin promoter/enhancer are expressed similarly in the heart of transgenic mice. However, the mole rat promoter/enhancer is at least than 10 times more active than the mouse promoter/enhancer in skeletal muscle and barely active in lens in the transgenic mice. The mole rat promoter does function during early embryonic in the transgenic mouse, but turns off later in development, suggesting the existence of a developmental switch for shsp/alphaB-crystallin gene expression. We concluded that the shsp/alphaB-crystallin promoter/enhancer underwent adaptive changes corresponding to the subterranean evolution of the blind mole rat. In FY2003 we provided evidence via transgenic mice that a mole rat DNA motif, possibly associated with Pax3-binding and located between the enhancer and proximal promoter, contributes to suppression of lens activity in the mole rat alphaB-crystallin promoter. In collaboration with Dr. Melinda K. Duncan (University of Delaware), we reported in FY2003 that the transcription factors Maf and Prox1 bind and activate the chicken betaB1-crystallin promoter by displacing Pax6, which represses promoter activity of this gene. This year we completed a research project on the involvement of Pax genes in jellyfish eyes. Jellyfish have complex eyes with a cellular lens, cornea and retina. Last year we showed that the jellyfish Pax protein, PaxB, is a structural hybrid of Pax6 (critical for eye development) and Pax2 (critical for eye and ear development), implicating a commonality between the eye and the ear in evolution. PaxB has a Pax2/5/8 DNA-binding paired domain and octapeptide and a Pax6 homeodomain. In addition to sequence similarities to Pax6 and Pax2/5/8, PaxB rescues a Pax2 eye mutation and can induce small ectopic eyes in the fly. PaxB activates the J3-crystallin promoter in transfection experiments, extending the fundamental role of Pax proteins for crystallin gene regulation throughout the animal kingdom. Thus, our data suggest that PaxB, rather than Pax6, might be the primordial gene for eye evolution.